Skip to content

Latest commit

 

History

History
41 lines (26 loc) · 1.04 KB

analysis.md

File metadata and controls

41 lines (26 loc) · 1.04 KB
layout title permalink
page
Analysis
/analysis/

Summary of Data

Characterization of Metrics

This plot shows the expectation and variance for datasets with n entities in the standard case where for a dataset with n entities, there are n ranking tasks for which the number of candidates is also n. The upper plots show the standard variants of the metrics (i.e., MR, GMR, HMR) and the lower plots show the respective reciprocal variants of the metrics (i.e., IMR, IGMR, MRR) as well as the Hits at K for all K up to half the order of magnitude of the number of entities. Closed form solutions were used when possible, otherwise 500 samples were used in a Monte-carlo simulation.

Other Plots

Arithmetic Mean Rank

Geometric Mean Rank

Inverse Harmonic Mean Rank (Mean Reciprocal Rank)

Hits at K